On 27 May 2017 3:30 a.m., "Bruce Kellett" <bhkell...@optusnet.com.au> wrote:
On 27/05/2017 11:46 am, David Nyman wrote: On 27 May 2017 at 01:44, Bruce Kellett < <bhkell...@optusnet.com.au> bhkell...@optusnet.com.au> wrote: > > I think it is the interpretation of the data that is theory-dependent. > Not at all. Data don't just sit there staring you in the face. What is data in terms of one theory is mere background noise in terms of another. That is what it means for the *interpretation* to be theory-dependent. Sure, that too, but prior to that you wouldn't necessarily be looking in that particular direction or with that particular interpretative mindset without having at least an implicit theory in the background. This insight led Popper to reject the notion of induction. As he (I believe correctly) pointed out, the very notion of the data on which induction putatively relies is theory-dependent and hence primarily deductive. Conjecture and refutation is a better account of how science (or any consistent reasoning) actually proceeds. I don't think that is an accurate account of Popper. It was the asymmetry between falsification and confirmation that lay at the heart of his rejection of induction. His account of scientific practice as "conjecture and refutation" was rather naive, and philosophy of science has long moved beyond this. In my view the dismissal of Popper is often based on a somewhat naive misreading. The process he described is rarely explicit, especially in the case of 'normal' science as distinct from attempts at a theoretical breakthrough. The asymmetry between falsification and confirmation is likewise frequently implicit rather than stated, but that doesn't mean that theories are ever more than provisional, even if our confidence in them grows the more they resist refutation. I need hardly quote you chapter and verse from the history of science. > But then you have a hierarchy of theories -- what is a new cutting-edge > theory today is tomorrow's instrument for data taking..... > > > > And the reductive aspect of theory is itself an implicitly ontological > commitment. > Not for the pragmatic instrumentalist. Even committed scientific realists > would only claim that it is only for our best, well-established, theories > that there is any suggestion that the suggested entities actually exist. > But we're not interested in "reifying" the ontology. It merely represents the unexplained part of an explanatory hierarchy. That is an unconventional definition of ontology. Perhaps one can say that ontology is theory-dependent in that any mature theory carries an implied ontology -- statistical mechanics implies an ontology in which atoms/molecules exist, the standard model of particle physics implies an ontology in which quarks and gluons exist -- but the theory itself does not rely on such an ontology. The elements of the theory may be nothing more than convenient fictions, as was for many years the status of quarks in the Gell-Mann quark model -- the predictive power of the theory would be in no way impaired. That's the sense in which it exists. It's the part that's "independent of us" simply because, although the basis of any explanation that follows, it doesn't itself rely on our explaining it. Quite. Atoms, quarks, and gluons, require no explanation in terms of the theory. They are just terms in the equations that the theory uses -- their existence or otherwise is not an issue for the success of the theory. So if a hierarchy of laws were to imply mutually inconsistent ontological commitments it would be to that extent incomplete and unsatisfactory. Indeed the holy grail of (Aristotelian?) science is a hierarchical "Theory of Everything" that is, in precisely this sense, ontologically consistent "bottom up all the way down", if you'll permit me a slogan of my own. The search for such a TOE has a chequered record in the history of science. Some still hope that such a theory is possible, but the negative induction from the past record would not lead one to be optimistic that any such theory exists or is possible. For these reasons I can't accept that your distinction between Platonic and Aristotelian modes of explanation has much real force. In practice, *any* effective mode of explanation must inexorably be constrained by its fundamental ontological commitments, That is the case only on your account of "explanation". If explanation does not rely on an underlying ontology, then it is not constrained by any such assumed ontology. Not all explanations need be reducible to your model of explanation. That's true of course Bruce, but I would think then that any such heterogeneous account of explanation is in serious danger of falling into inconsistency. Why? Because their respective ontological commitments may turn out to be in conflict. This crops up all the time in the attempt to unify knowledge over time. So long as the respective fields can rub along despite this it can be tolerated. But at the point where such inconsistencies present a serious inconvenience there is usually an attempt to reconcile the ontological commitments of previously disparate theories. And so I can't agree that "my account" of a mutually-consistent reductive hierarchy of explanation is substantially different from what would generally be accepted, if only implicitly, as the ultimate aim of mechanistic explanation tout court. Whatever you say, it must be the case, in the final analysis, that the entire hierarchy of explanation must ultimately be reducible to a common set of ontological commitments or else risk frank inconsistency between its constituent elements. Not at all. There is no inconsistency between explaining a reduction in infection in terms of an antibiotic injection, and an explanation in terms of the fundamental biochemistry of bacteria. Since that is so obviously the case, I might have hoped that you would have given me the benefit of the doubt that something of this sort was anything to do with my point. To make use of your example, however, it would certainly be inconsistent if the science behind antibiotics was based on a theory, say, of djinns rather than one based on the fundamental biochemistry of bacteria. There would thereby be a greater likelihood of divergence between prediction and outcome. I should re-emphasise here that by ontology I mean merely those explanatory entities and relations that are in themselves not further explicated within the theory, but rather serve as the explanatory point of departure. Of course, you are perfectly correct that, pragmatically, no such fully-reducible hierarchy has yet been achieved. Indeed it is more than possible that it never will be. But I think it would nonetheless be an evasion to claim that ontological consistency isn't in a rather strong sense an underlying assumption of the scientific method itself. Only for the scientific realist, and then the ontology, while assumed to be consistent, is not necessarily accessible. The realism in question concerns explanation not reification. The latter is entirely beside the point. All modern science is implicitly based, as Bruno correctly points out, on a common assumption of mechanism, variously realised depending on whatever level of rigour and detail is achievable at any present state of the art (if you'll forgive the term). And that mechanistic assumption is in turn based implicitly on its reducibility to a "fundamental" or irreducible level (explanatorily speaking). This is true even in the case that any currently assumed such level is provisional, as indeed must always be the case, because we are speaking of explanatory consistency, not ultimate truth. Remember that the assumption of an arithmetical ontology in computationalism is likewise not presented as an ultimate truth but rather as the provisional basis for explanatory consistency in what follows. However, all that said, if your views are indeed as set out above, it's hardly surprising that you don't feel very compelled by the explanatory approach of computationalism, in which mutual consistency between ontological and epistemological components at all levels of explanation is indispensable. Too bad. Too bad for whom? Too bad for the conversation. It seems that we can rarely get very far from an argument over theoretical ground rules. And yes, I am not compelled by the idea that explanation must be rooted in an underlying ontology. Hmm.. But as I've said above, the whole explanatory schema of computationalism depends on it. So unless you can provisionally adopt that mindset to see where it might take you we are forever stuck in "talks about talks". I don't believe one need feel that that one's whole philosophy of science is at stake in such an experiment, surely? David Bruce -- You received this message because you are subscribed to the Google Groups "Everything List" group. To unsubscribe from this group and stop receiving emails from it, send an email to everything-list+unsubscr...@googlegroups.com. To post to this group, send email to everything-list@googlegroups.com. Visit this group at https://groups.google.com/group/everything-list. For more options, visit https://groups.google.com/d/optout. -- You received this message because you are subscribed to the Google Groups "Everything List" group. To unsubscribe from this group and stop receiving emails from it, send an email to everything-list+unsubscr...@googlegroups.com. To post to this group, send email to everything-list@googlegroups.com. Visit this group at https://groups.google.com/group/everything-list. For more options, visit https://groups.google.com/d/optout.